US 11402453

I'll verify details against live sources before summarizing.

Let me check for the actual claim set and any litigation specific to this patent.

US 11,402,453 B2 — Verification Summary

Note on scope: I searched for this exact number (US 11,402,453 / US11402453B2 / application 15/723,242) and did not return results for look-alike numbers. The bibliographic data below is corroborated across Google Patents, PubChem (NCBI patent record), and Justia. The one caveat is flagged in §3.

1. Bibliographic data

Sources: https://patents.google.com/patent/US11402453B2/en ; https://pubchem.ncbi.nlm.nih.gov/patent/US-[11402453](/patent/11402453)-B2 ; https://patents.justia.com/patent/11402453

2. Abstract (verbatim from the authoritative record)

"A method is for using measurement data of an object of examination for a post-processing process. In an embodiment, the method includes recording first measurement data, the first measurement data being previously determined via a medical imaging modality; automatically analyzing the first measurement data based on defined criteria and automatically inspecting a set of control parameters with aid of an analysis of the first measurement data using defined criteria with regard to second measurement data, the second measurement data being previously recorded via the modality using the set of control parameters, wherein the defined criteria include at least one of a post-processing capacity and identification of at least one image characteristic; and using at least one of the first measurement data and the second measurement data in a post-processing process. A control device and a medical imaging system are also disclosed."

3. Plain-language overview of the independent claims

Important uncertainty flag: the full text I have access to contains the abstract, summary, and detailed description, but not the granted claim set verbatim. The overviews below are drawn from the "SUMMARY" section of the specification (which restates claim-category subject matter) rather than from literal granted claim language. I will not fabricate claim numbers or exact wording.

Based on that summary, the patent appears to have at least three independent claims in three categories:

• Independent method claim — "method for using measurement data of an object of examination for a post-processing process." Steps: (1) record first measurement data that was previously determined via a medical imaging modality (may be topogram data, raw data, or image data — it need not support 3-D reconstruction); (2) automatically analyze that first measurement data against defined criteria, and also automatically inspect a set of control parameters (the acquisition protocol) by analyzing the first measurement data with regard to second measurement data that would be recorded using that parameter set; (3) the "defined criteria" comprise a post-processing capacity of the measurement data and/or identification of at least one image characteristic; (4) optionally modify the set of control parameters; (5) optionally record second measurement data using the (possibly modified) parameter set; and (6) use the first and/or second measurement data in the post-processing process. The point of novelty is doing this before post-processing, using criteria that go beyond simple computable metrics such as SNR/CNR.

• Independent control-device claim. As recited in the summary: a control device comprising an acquisition unit to record first and second measurement data determined via a medical imaging modality; a memory storing computer-readable instructions; and a processor configured to execute those instructions to automatically analyze the first measurement data with regard to post-processing capacity based on defined criteria and to automatically inspect a set of control parameters with the aid of that analysis, and to use the first measurement data in the post-processing process. (The specification's generic device recitation also names an analysis unit, a post-processing unit, and optionally a modification unit, connected via a bus in the FIG. 3 embodiment.)

• Independent medical-imaging-system claim. A system comprising the above control device plus a medical imaging modality (CT device, MRT/MRI device, tomosynthesis device, ultrasound device, or angiography unit).

• A non-transitory computer program product claim category is also described (computer program loadable into a control device's storage, with control sections to perform the method). Whether this is an independent claim or handled only in the description is not confirmable from the material I have.

Dependent/embodiment features described (likely mirroring dependent claims): machine-learning-based analysis using a database of reference examinations; a training method where a discriminator (e.g., a neural network) is trained on post-processed learning data judged "worked / did not work," including "in-line" retraining during operation; image characteristics including obesity, lesions, embolisms, motion artifacts, metal artifacts; a plausibility check of the identified image characteristic using demographic data or patient-file information; combining first and second measurement data (e.g., linear combination with coefficients set by the analysis results); control parameters including tube voltage, tube current, filtering, dual energy, reconstruction method, layer thickness, triggering, gantry tilt, pulse sequence, and/or delay; iterative repetition until a termination criterion (e.g., sufficient post-processing capacity or a radiation-exposure limit); and real-time imaging operation.

Sources: Google Patents full text, https://patents.google.com/patent/US11402453B2/en

4. Litigation / CAFC docket search (as requested)

I found no USPTO proceeding (IPR/PGR), no U.S. district court case, and no Federal Circuit 2026 docket involving US 11,402,453. Specific notes on what the search did surface, so the negative result is unambiguous:

• The only 2026 Siemens Healthineers patent litigation prominently reported is Hologic, Inc. v. Siemens Healthineers AG et al., UPC CFI Local Division Düsseldorf, Case Nos. UPC_CFI_758/2024 and UPC_CFI_259/2025, concerning EP 2 352 431 B1 ("method and system for controlling x-ray focal spot characteristics for tomosynthesis and mammography imaging"). Judgment issued 10 June 2026 (injunction, recall, dismissal of Siemens' revocation counterclaim). This is a different patent and a different technology — it is not US 11,402,453. Sources: https://www.unifiedpatentcourt.org/sites/default/files/files/api_order/2026-03-09%20R36%20Order.pdf ; https://www.eplaw.org/blog/detail/upc-hologic-v-siemens-healthineers/ ; https://pt.dotmed.com/news/story/66479
• A search hit referring to a "'453 decision" and "SSBG '453 claims" concerns an unrelated Federal Circuit/SSBG telecommunications case (ET CI patents), not US 11,402,453. I am flagging it so it is not mistaken for a hit on this patent.
• I could not directly query the CAFC docket system or PACER/PTAB APIs with the tools available; the negative litigation result is based on targeted web searches and should be treated as "no reported activity found," not as a certified docket clearance.

5. Residual uncertainties

• Claim text not verified verbatim. The claim set of US 11,402,453 was not present in the authoritative record I retrieved, and my searches did not return a claim listing. Exact independent-claim wording, claim count, and claim numbering should be confirmed against the USPTO PatentCenter/PatFT record or the granted-PDF claims before being quoted.
• Title wording. The Google Patents page header, the abstract, and Justia all use "Method and system for determining sufficiency of measurement data for post-processing process," while the specification's own FIELD/SUMMARY language repeatedly says "method for using measurement data… for a post-processing process." Both phrasings appear in the record; the DB title is the "determining sufficiency" version.
• Expiration date (2037-11-27) is Google Patents' own annotation (reflecting any PTA), not a legal conclusion.
• Assignee chain (Siemens Healthcare GmbH → Siemens Healthineers AG) is per Google Patents' reassignment entries dated 2023-12-20.

I'll search for litigation involving US patent 11402453 across the sources you mentioned.

Let me search more targeted litigation databases.

Let me run additional targeted searches to confirm whether any litigation or PTAB proceedings exist.

Summary of Findings — Litigation Involving US Patent 11,402,453

No litigation or post-grant proceedings involving U.S. Patent No. 11,402,453 were found.

I searched litigation-focused sources (Unified Patents litigation resources/caselist, PACER-derived records, CourtListener/RECAP, USPTO PTAB/PTACTS documents, and general web search) for the exact identifier 11402453 and for the patent paired with its parties (Siemens Healthineers / Siemens Healthcare GmbH) and inventors (Grzegorz Soza, Sebastian Schmidt). I could not identify a single district-court case, ITC action, PTAB (IPR/PGR) proceeding, or CAFC appeal in which US 11,402,453 has been asserted or challenged.

What the searches did return (all distinguishable — NOT this patent)

To be transparent about the "hits" that appeared, none involve US 11,402,453:

• UK Companies House registration No. 11402453 — "J McKinnie Electrical Limited," a UK private company. Unrelated to any US patent. (companiesintheuk.co.uk)
• Kenyan national/ID number 11402453 — appearing in an IEBC electoral register PDF. Unrelated.
• PTAB documents (Samsung, Unified Patents, MasterObjects, Bradium, etc.) — these cite the un-related patents/numbers in other proceedings; a search-engine match on a different patent number, not on 11,402,453.
• Hologic, Inc. v. Siemens Healthineers AG et al. — UPC Düsseldorf Local Division, Case Nos. UPC_CFI_758/2024 and UPC_CFI_259/2025, concerning European Patent EP 2 352 431 B1 (mammography / "Flying Focal Spot"). This is a different patent (EP, not US), and a different technology. It is a real Siemens litigation but not US 11,402,453.

Caveats and confidence

• Confidence: Moderate-to-high that no litigation currently exists, but I cannot claim absolute certainty. Free/public searches may miss recently filed complaints, sealed filings, or cases docketed only in PACER (which I could not query directly here), and non-US actions may not surface.
• US 11,402,453 (granted Aug. 2, 2022; priority Oct. 12, 2016; assignee Siemens Healthineers AG) is a diagnostic-imaging data-processing patent. Notably, its grant date is relatively recent and its subject matter is internal imaging workflow/control — characteristics that make it less likely to be an NPE assertion target and consistent with the absence of observed litigation.
• Recommended verification steps if a definitive answer is required: run a direct party/patent query in Docket Navigator, Lex Machina, Unified Patents Portal (portal.unifiedpatents.com), and USPTO Patent Center / PTAB using the exact application number 15/723,242 (the application that issued as 11,402,453), plus a PACER docket search.

Bottom line: Based on the available information, there is no known litigation involving US Patent 11,402,453. I did not find any case to report plaintiff, defendant, jurisdiction, case number, filing date, or status for — and I have deliberately excluded near-matches (e.g., the Hologic v. Siemens UPC matter and unrelated IDs) so as not to misattribute activity to this patent.

I'll verify the structured data with targeted searches before writing up.

Proceedings overview

Zero AIA trial proceedings are on file against US 11,402,453. The structured PTAB block supplied in the prompt is the canonical source and reports no AIA trial proceedings as of the most recent ODP ingest; my independent web checks (Google Patents, PubChem patent record, PTAB/PTABTS document search, Docket Alarm PTAB case index, and CAFC/CourtListener-adjacent results) surfaced no IPR, PGR, or CBM petition, no institution decision, no Final Written Decision, and no Federal Circuit appeal involving this patent. Breakdown therefore stands at: 0 active / 0 invalidated / 0 sustained / 0 settled / 0 institution-denied.

Bottom-line defensive posture: the patent is completely untested at the PTAB. There is no cancellation to hand you a free walk, and — equally — no adverse FWD or estoppel ruling that has hardened it. A defendant today faces an unfought patent, which means you get to write the invalidity record from scratch, choose your own art, and pick your own forum (IPR vs. district court vs. both) without inheriting anyone else's bad briefing.

Per-proceeding detail

There are no proceedings to iterate. I am not going to manufacture proceeding numbers, panels, or FWDs to fill the template — with a count of zero, every field in the per-proceeding template would be a fabrication. The absence of PTAB activity is the substantive finding.

Signals I checked and rejected as false positives (so a defendant doesn't chase them):

• Hologic, Inc. v. Siemens Healthineers AG et al., UPC CFI_758/2024 and CFI_259/2025 (Düsseldorf Local Division, decision 2026-03-10 per the reporting, panel incl. judges Thom, Rinken, Rinkinen and technically qualified judge Kitchen) concerns EP 2 352 431 B1 — Hologic's X-ray focal-spot control patent for breast tomosynthesis — not US 11,402,453. Siemens is the defendant there, and the patent owner is Hologic. Different patent, different family, different party posture.
• Various PTAB hit snippets mentioning "Siemens" (e.g. IPR2022-00511 Boston Scientific v. Nevro; IPR2025-00943 Tesla v. Granite Vehicle Ventures; IPR2025-01252 Samsung v. Omni MedSci) are unrelated patents whose exhibits or briefing merely quote Siemens/Siemens-adjacent file histories.

Strategic summary

Claim status. Every claim of US 11,402,453 is UNTESTED — none canceled, none sustained, none construed by the Board. I do not have the issued claim text or claim count in the authoritative material supplied, and I am not going to state a claim count from memory. Practically: the independent claims you will have to attack are the ones recited in the granted patent, and no claim has yet been through an institution-stage merits screen. That cuts both ways — the claims have never been narrowed by an adverse PTAB construction or a certificate of correction, so you should read them at the full breadth a district court would apply under Phillips, not the narrower reading BRP-era IPR panels sometimes imposed.

Estoppel landscape. § 315(e)(2) estoppel is a blank slate. No petitioner exists, so nobody is statutorily barred from raising art they raised or reasonably could have raised in a prior IPR. If you file first, you own the proceeding and you alone bear the estoppel risk going forward. If you are already in district court, note that there is no IPR clock running and no § 315(b) one-year bar problem unless and until you are served with a complaint on this patent — if you have been served, count the year carefully, because the 2017-10-03 filing date means this is a post-AIA patent and IPR is squarely available (not a pre-AIA-only or first-inventor-to-file carve-out situation). Also worth modeling: because the claims recite a computer-implemented workflow (automatic analysis of measurement data against defined criteria, inspection of a set of control parameters, selective re-acquisition), a § 101 Alice/Mayo challenge is a live non-IPR ground that the Board cannot hear — keep it in district court, where the estoppel of a filed IPR won't reach it.

Pattern signals. No repeat-petitioner pattern (impossible with zero petitions). No indication of a defensive aggregator (Unified Patents, RPX, or similar) in the chain — the assignee chain is clean Siemens corporate: filed 2017-10-03 by Siemens Healthcare GmbH, assignment recorded 2018-01-02 (inventors Soza and Schmidt), granted 2022-08-02, and reassigned to Siemens Healthineers AG on 2023-12-20, with an adjusted expiration of 2037-11-27. The patent owner has never had to defend a PTAB challenge and therefore has no appellate track record on this patent to learn from.

Recommended next steps

1. Do not expect a silver bullet. There is no FWD to link, no canceled claim to quote, and no institution decision to distinguish. Any advice along the lines of "claims 1-5 are dead" is wrong for this patent. The patent stands as granted.
2. Because there's no PTAB record, the district court record is your whole case. If you have been served, run a standard pre-suit review: (a) an invalidity search against the 2016-10-12 priority date (DE 102016219887.4) covering CT/MR acquisition-protocol adaptation and image-quality-driven re-scan art; (b) a § 101 mapping of the "automatic analysis … defined criteria … post-processing capacity" steps onto Alice step two; (c) claim-chart the "post-processing capacity" and "image characteristic" limitations, which are functional and may be vulnerable to § 112(b) indefiniteness or written-description attack if they are claimed only as a result to be achieved.
3. Watch the family, not just this patent. Siemens Healthineers is an active enforcer/defendant in the UPC (see the Hologic dispute above, involving a different EP), so litigation activity around Siemens imaging patents is occurring elsewhere. If a demand arrives citing US 11,402,453, check for continuations, divisionals, or sibling US patents in the same German priority family — assertion often migrates to the sibling once one patent takes an IPR hit, and here no such hit exists yet.
4. If you file, file early. With no § 315(b) clock triggered by prior service and no competing petitioner, you have the rare luxury of forum and timing choice. Institution-stage milestones, if you file: Patent Owner Preliminary Response due ~3 months after notice; institution decision due within 6 months of the petition; and a statutory 1-year trial deadline running from institution under § 316(a)(11) for the FWD (§ 316(a)(11)) — with the Board's usual practice of setting the oral hearing roughly 2–3 months before the FWD deadline.

Verification hook. I confirmed no proceedings on the canonical ODP list and none via secondary public sources. If you want a belt-and-suspenders check before relying on this, query PTAB E2E / the USPTO Patent Trial and Appeal Board End-to-End system directly for "11,402,453" and the application number 15/723,242 — I could not retrieve a live E2E result in this session, so I am reporting "no PTAB activity found" rather than "E2E affirmatively returned zero hits."

I'll research this patent's ownership chain using the available sources.

Let me dig deeper on the specific assignment records and reel/frame data.

Ownership Analysis — US 11,402,453 B2

Patent: Method and system for determining sufficiency of measurement data for post-processing process
Application: US 15/723,242 · Priority: DE 102016219887.4 (2016-10-12) · Filed: 2017-10-03 · Granted: 2022-08-02

Data-integrity note up front: I was able to confirm the events and parties in this chain from Google Patents legal events and corroborating public sources, but I could not retrieve the USPTO reel/frame numbers or the recording correspondent for either assignment in this session — Assignment Center was not reachable through the search tooling available to me. Everything below that is sourced is labeled as sourced; anything I could not verify (reel/frame, correspondent) is explicitly marked unverified rather than reconstructed. Per the operating rules, I will not fabricate reel/frame or correspondent data.

Inventors

Both named inventors are Siemens medical-imaging personnel; the subject matter (CT/MRI acquisition-parameter control, artifact detection, machine-learning-based post-processing sufficiency) is squarely within the Siemens Healthineers Computed Tomography / Advanced Therapies engineering organization, and both names recur on other Siemens Healthineers CT/AI filings (e.g., the DE-published family document listing SCHMIDT, Sebastian as inventor). Employer attribution is therefore determinable with reasonable confidence from the family and inventor-name overlap, though I did not independently pull a payroll-period employment record.

Unusual-pattern check — no flag. There is no evidence either inventor left Siemens Healthineers within 12 months of the 2017-10-03 filing. Both remain associated with Siemens Healthineers imaging work in later filings in the same family. This is not a pre-fire-sale departure pattern.

Original assignee

Siemens Healthcare GmbH (Erlangen, Germany) — the assignee named on the issued patent and the original assignee of record.

• Primary line of business: The Siemens group's medical-technology operating entity — CT, MRI, X-ray, ultrasound, molecular imaging, and the associated software/algorithm portfolio (this patent's post-processing metrics — ctFFR, bone removal, volumetry, CAD, ML-based image-sufficiency analysis — are exactly that portfolio).
• Product embodiment: Yes. This is not a paper patent. Siemens Healthineers ships CT scanners (SOMATOM) and MRI systems (MAGNETOM) plus post-processing/analytics software (e.g., syngo.via, AI-Rad Companion) that practice medical-image post-processing of precisely the kind claimed. The claims read on the vendor's own scanner-side acquisition-parameter control and post-processing pipeline.
• Current status: Operating. Siemens Healthcare GmbH was the German operating vehicle at spin-off; Siemens Healthineers AG listed in Frankfurt (SHL) in 2018, and the group announced consolidation of the German operating business from Siemens Healthcare GmbH into Siemens Healthineers AG (a so-called "lift and drop"). Not dissolved, not in bankruptcy. No Chapter 7/11 event.

Assignment timeline

The chain is short and consists of exactly two recorded transfers: (1) the confirmatory employment assignment from the inventors to the employer, and (2) the internal corporate transfer of the portfolio to the post-IPO parent.

• Recorded 2018-01-02 (execution date not separately confirmed; recorded roughly three months after filing) — Reel/Frame unverified

  • Conveyance: Assignment of Assignors' Interest (inventor-to-employer)
  • Assignor: Grzegorz Soza; Sebastian Schmidt
  • Assignee: Siemens Healthcare GmbH
  • Correspondent: unverified (not retrievable this session)
  • Context: Standard employment/confirmatory assignment of inventors' rights — the inventors' rights vest in the employer. No third-party or licensing entity involved.

• 2023-12-20 (executed and recorded) — Reel/Frame unverified

  • Conveyance: Assignment of Assignor's Interest
  • Assignor: Siemens Healthcare GmbH
  • Assignee: Siemens Healthineers AG
  • Correspondent: unverified (not retrievable this session)
  • Context: Internal corporate reorganization — transfer of the portfolio to the listed parent, consistent with the publicly announced consolidation of the German operating business into Siemens Healthineers AG (announced 2023; corroborated by the group's 2023 restructuring communications). This is an intra-group transfer, not a sale.

If Assignment Center shows only these two events when queried directly, that is the complete chain. I am not able to state a reel/frame for either entry, and no correspondent has been confirmed.

Timeline diagram

timeline
    title Ownership of US 11402453
    2016 : German priority application filed
    2017 : US application filed by Siemens Healthcare GmbH
    2018 : Inventors assign rights to Siemens Healthcare GmbH
         : US application published
    2022 : Patent granted as US 11402453 B2
    2023 : Intra-group transfer to Siemens Healthineers AG

NPE / troll-pattern signals

1. Shell-entity transfer — not present. No transfer to any entity bearing "IP / Patents / Licensing / Holdings / Ventures." The only post-inventor transfer (2023-12-20) moved the patent up to the publicly listed parent, Siemens Healthineers AG (Frankfurt: SHL) — a large operating company, not a single-purpose LLC. No registered-agent-service address is implicated.

2. Known asserter in the chain — not present. Neither assignee (Siemens Healthcare GmbH, Siemens Healthineers AG) appears on any public NPE list (Acacia, Marathon, Intellectual Ventures, IPNav, Wi-LAN, Conversant/Mosaid, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation, Spangenberg entities). No Unified Patents / RPX high-frequency-plaintiff match. No litigation naming this patent surfaced.

3. Repeat correspondent across the chain — unclear. I could not retrieve the correspondent of record for either assignment, so I cannot test for recurrence. Cannot call this signal either way. (Note: large-cap corporate patent departments routinely use a single outside firm for all domestic and foreign portfolio recordings, so a recurring corporate correspondent would not by itself be an NPE tell — the signal only matters against a chain of unrelated shell LLCs.)

4. Cascading transfers — not present. Two recorded events over ~6 years, both between related Siemens entities, does not approach the "<24 months through chained LLCs" pattern. The 2023 transfer moved to the parent, the opposite of a cascade into shells.

5. Pre-litigation transfer — not present. No infringement suit naming this patent was found, so there is no litigation date against which a transfer could be timed. The 2023-12-20 transfer is separated from grant (2022-08-02) by ~16 months and is explained by corporate restructuring, not assertion.

6. Bankruptcy fire-sale — not present. No Chapter 7/11 proceeding involving Siemens Healthcare GmbH or Siemens Healthineers AG; no evidence this patent was sold in a bankruptcy estate.

7. Privateering — not present. No transfer to an NPE asserting on Siemens' behalf; the patent remains inside the Siemens Healthineers group that makes the products.

8. Defensive aggregator (anti-NPE) — not present. The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. It terminates at the original operating company's listed parent.

Verdict

Defensive / non-asserting (operating-company hold).

Definitional caveat, stated honestly: this patent does not terminate at a defensive aggregator, so it is not "defensive" in the strict RPX/AST/LOT sense. It is placed in the non-NPE bucket because that is the only correct answer for the evidence: the chain contains no NPE signals whatsoever. The only two recorded events are (1) the inventor-to-employer assignment recorded 2018-01-02 (Soza and Schmidt → Siemens Healthcare GmbH) and (2) the intra-group transfer recorded 2023-12-20 (Siemens Healthcare GmbH → Siemens Healthineers AG), both between related Siemens operating entities, with no litigation and no shell LLC anywhere. The current holder, Siemens Healthineers AG, is a €23B-revenue operating med-tech company that ships CT/MRI systems and post-processing software embodying the claims — the opposite of a licensing-only asserter.

Verification link: search patent number 11402453 at the USPTO Assignment Center — https://assignmentcenter.uspto.gov/ (mirror: https://assignment.uspto.gov/patent/index.html). Confirm the reel/frame and correspondent fields directly, which I could not retrieve here.

Two residual gaps in this analysis (both flagged, neither fabricated): (i) reel/frame numbers for both assignments are unverified; (ii) the recording correspondent(s) of record are unverified, so signal #3 could not be tested. If you can run the Assignment Center query, supplying those fields would close both gaps — though neither is likely to change the verdict given the absence of any third-party or shell-entity party in the chain.

I'll research US patent 11402453 and its prior art citations.

Let me dig deeper into the specific citation lists for this patent.

Let me try to retrieve the actual citation table from Google Patents for this specific patent.

I'll research this patent's claims, cited prior art, and the relevant technological landscape before giving the § 103 analysis.

Let me pull the full claim set and identify the prior art cited on this patent's page.

§ 103 Obviousness Analysis — US 11,402,453 B2

Patent: US 11,402,453 B2 ("Method and system for determining sufficiency of measurement data for post-processing process")
Inventors: Grzegorz Soza; Sebastian Schmidt | Assignee: Siemens Healthineers AG (orig. Siemens Healthcare GmbH)
Priority: DE 102016219887.4, filed 2016‑10‑12 | US filing: 2017‑10‑03 | Granted: 2022‑08‑02
Governing law: AIA §§ 102/103 (effective filing date after 2013‑03‑16)

1. Threshold note on the record supplied

I have to flag a limitation on the evidence base before giving the analysis, because it materially affects confidence.

The Google Patents page text supplied in the task does not contain a "References Cited" or "Cited By" (Prior Art) table. That section of the Google Patents page is rendered dynamically and was not captured in the fetched text. What is present in the record are only the "Prior art keywords" field (measurement data; post-processing process; data; desired post) and the specification's own admissions of prior art.

Consequently, I am working from two sources:

1. Verified citations recovered by search (one clearly relevant Siemens reference — see below), and
2. The patent's own Background section, which is itself a binding admission of prior art and scope.

Where I cannot verify a reference number from this page's prior-art section, I say so rather than inventing one. That is the correct professional posture here; a § 103 chart built on unverified identifiers would be worthless.

2. The claims to be analyzed

Claim 1 is a method claim. From the granted text and the dependent-claim cross- references (verified via Justia and Google Patents), claim 1 requires, in substance:

(a) recording first measurement data previously determined via a medical imaging modality;
(b) automatically analyzing the first measurement data based on defined criteria — where the analyzing is performed using a neural network, and where the claim requires determining whether the first measurement data is insufficient with regard to a desired post-processing process among a plurality of post-processing processes;
(c) the desired post-processing process including a data modification operation or a data analysis operation;
(d) the defined criteria including at least one of (i) a post-processing capacity of the desired post-processing process or (ii) identification of at least one image characteristic associated with the post-processing capacity;
(e) the neural network being trained on reference measurement data of reference objects, including positive reference measurement data determined to provide a sufficient result image when processed by the desired post-processing process;
(f) automatically inspecting a set of control parameters with the aid of the analysis, with regard to second measurement data recorded via the modality using that set;
(g) obtaining a modified set of control parameters;
(h) recording second measurement data using the modified set; and
(i) using the first and/or second measurement data in the post-processing process.

Dependent claims add: plausibility check of the image-characteristic identification (cl. 6); the object of examination remains in the modality until analysis is complete (cl. 8) or until the second measurement data is obtained (cl. 18); the control-parameter list — tube voltage, tube current, filtering, dual energy, reconstruction method, layer thickness, triggering, gantry tilt, pulse sequence, delay (cl. 9); iterative repetition to a termination criterion (cl. 10) and real-time imaging (cl. 11–12); image data/raw data/topogram data (cl. 13–14); user selection of the post-processing process (cl. 19); negative reference measurement data (cl. 22); and linking result images to learning data by outcome (cl. 23). Claim 25 is the control-device counterpart; claims 5/15/16 are the CRM and computer-program-product claims.

3. Verified prior art on the record

R1 — US 2007/0008172 A1 (Siemens), "Post-processing of medical measurement data," pub. 2007‑01‑04

Verified via Google Patents, FreePatentsOnline and Patents-Review. Clearly § 102(a)(1) art against a 2016 priority date.

R1 discloses a method for automatically selecting a post-processing method for medical measurement data, comprising:

• registering post-processing components so that "it is defined for each post-processing component what type of measurement data are required by it";
• acquiring and/or deriving context data with respect to the measurement data, drawn from an acquisition context, a procedure context, an observation context and/or a post-processing context (the acquisition context expressly including "the type of acquisition device, the age, other systems connected etc.");
• parsing the measurement data enriched with the context data according to its format;
• evaluating the parsed data in dependence on the registered post-processing components so that "an optimally designed post-processing method can be selected for the respective measurement data"; and
• selecting at least one optimally designed post-processing method.

R1 expressly frames the problem in terms identical to the '453 patent's field: "The selection of an unsuitable post-processing tool can thus completely falsify the measurement data originally acquired correctly and render them useless overall."

Mapping to claim 1: R1 discloses (a) recording measurement data; (b) automatic (machine-executed) analysis against criteria defined by registered post-processing requirements; (d)(i) an assessment of whether the data meets the input requirements of a desired post-processing process — i.e., a "post-processing capacity" determination, though R1 expresses it as a data-type/format/content match rather than an image-quality score; (c) post-processing operations of both the modification and analysis type.

What R1 does not disclose: a neural network; the specific "insufficient" determination framing; image-characteristic identification (artifact/obesity/lesion classes); any closed loop back to acquisition — i.e., modifying the set of control parameters and re-acquiring second measurement data; and training labels derived from post-processing outcomes.

R2 — US 8,036,434 B2 (Siemens), "Post-processing of medical measurement data"

Verified in substance from the patent PDF text; moderate confidence it is the granted counterpart of R1 (same disclosure — DE 103 47 433 citation, same "selecting at least one post-processing method" object and claim structure). I flag this because I could not confirm the family relationship from the page itself.

R2 adds the explicit claim-level statement of the same subject matter: "a method for selecting at least one post-processing method for the post-processing of measurement data which are present in a predetermined form," with registration of components, context-data derivation, parsing, evaluation and selection.

R1 and R2 are, for § 103 purposes, cumulative. One of them supplies the teaching; treat them as a single evidentiary unit unless the family relationship is confirmed.

R3 — The patent's own admitted prior art (Background, verified from the supplied text)

The '453 specification admits the existence of:

• CAD (computer aided detection) algorithms "to find lesions" — i.e., automated identification of image characteristics (d)(ii);
• artifact-caused unsuitability for post-processing (motion artifacts, obesity, metal implants, stents);
• operators' visual inspection of image suitability as the then-existing practice;
• post-processing requiring higher dose or added contrast agent, and the clinical consequences of discovering unsuitability after the patient has left;
• SNR, CNR and image-sharpness as conventional computable metrics.

Admissions in the specification are prior art and are usable in an obviousness challenge.

4. The obviousness analysis

4.1 Legal framework applied

Under Graham v. John Deere and KSR Int'l v. Teleflex, I consider scope/content of the prior art, differences from the claims, PHOSITA level, and secondary considerations. The critical question is not whether any single reference teaches everything, but whether the differences would have been obvious to a POSITA with a rational underpinning (KSR; In re Kahn), including combinations of known elements "according to known methods" to yield predictable results, and combinations that are "obvious to try" from a finite number of identified, predictable solutions.

PHOSITA: a medical-imaging engineer/scientist with a graduate degree (or equivalent experience) in biomedical engineering, medical physics or computer science, with ≥3 years' experience in CT/MR image processing and ≥1 year in machine-learning-based medical image analysis as of 2016‑10‑12.

4.2 Ground 1 — R1 (or R2) in view of machine-learning image-quality/artifact-classification art and adaptive scan-parameter art

Primary reference: R1 / R2.
Secondary references (category-level, see caveat below):

• R4 — the well-established body of CNN/deep-learning medical image quality and artifact detection/classification art existing well before the 2016 priority date, including machine-learning classification of CT/MR images by artifact type and severity.
• R5 — the well-established adaptive scan-parameter control art: automatic exposure control / automatic tube-current and tube-voltage modulation (Anatomy-based or topogram-based), automatic protocol selection, bolus triggering, ECG-gated and arrhythmia-rejection re-scan, motion-correction re-scan, and iterative reacquisition while the patient remains on the table.
• R6 — CAD for lesion/embolism detection, which the '453 specification itself admits as prior art.

Caveat, stated plainly: I verified R1/R2/R6 by name. I could not verify specific patent numbers for R4/R5 from this page's prior-art section, because that section was not captured. R4 and R5 are cited here as established technology categories whose pre-2016 existence in the art is, in my assessment, beyond reasonable dispute (and is corroborated by the patent's own admission of CAD, and by the later Philips and Siemens filings that cite the '453 family — see §4.5). A § 103 chart for litigation would require naming and mapping specific references; I cannot responsibly do that from this record.

Element-by-element:

The critical motivation argument for the closed loop (g)/(h). This is the only element with no close counterpart in R1, and it is where the obviousness case must be won or lost. The rational underpinnings are:

1. Same field of endeavor, same problem, same solution type. R1 addresses exactly the problem the '453 patent addresses — post-processed results being made useless by mismatched input data. R5 addresses exactly the problem of acquisition parameters that produce inadequate data. Combining a detection stage with a correction stage is the paradigm of a feedback control loop, a combination KSR expressly blesses.
2. The art's own time pressure supplies the incentive. The '453 Background admits the clinical problem: unsuitability is discovered "when the patient is no longer in the modality," requiring recall and a completely new acquisition. R1 does not solve that; it solves the narrower problem of picking a better tool post hoc. A POSITA reading R1 in 2016, facing the admitted recall problem and knowing that re-scan-while-on-table was routine practice, would have had a strong, articulated reason to add the reacquisition branch.
3. Market/competitive forces — scanner makers competed on dose efficiency and first-pass diagnostic yield; an automatic "is this scan usable for ctFFR / perfusion / volumetry, and if not, re-shoot it now" feature is a design incentive KSR recognizes.
4. Finite, predictable solutions. Only a small number of responses to a "data insufficient" determination exist: use it anyway, use a different post-processing process, or re-acquire with changed parameters. The claim covers the third. KSR "obvious to try."
5. The control-parameter list of claim 9 is entirely conventional. Tube voltage, tube current, filtering, dual energy, reconstruction method, layer thickness, triggering, gantry tilt, pulse sequence and delay are the standard knob set of every CT/MR protocol; selecting among them to improve image quality is routine optimization of known parameters — the classic case of predictable variation.

4.3 Ground 2 — the "patient remains in the modality" and real-time claims (cl. 8, 11, 12, 18, 10)

Claims 8, 11, 12, 18 (and 10) require the object of examination to remain in the modality until analysis/obtaining second data, and iterative repetition/real-time imaging. These are enablement-by-practice limitations rather than technical advances: they recite the logistics of performing Ground 1's loop while the patient is still on the table. The patent treats these as advantages, not as structure. In view of R1+R4+R5, and of known real-time/fluoroscopic reconstruction and immediate-feedback re-scan capability in CT, these claims would be obvious.

For the topogram-based embodiment (cl. 13/14, and Figure 8), R1's context data plus the well-established topogram/scout-driven protocol-setting art supply the teaching. The specification itself concedes that topogram data "is not suitable for three-dimensional post-processing," so the claim's contribution reduces to: recognize from the scout image that a full scan is needed, then change the protocol and acquire it. That is what every topogram is for.

4.4 Ground 3 — Claim 6 (plausibility check)

Claim 6 requires a plausibility check of the image-characteristic identification — i.e., a confidence/probability determination, with comparison against demographic information (age-disease correlation) or patient-file data (implant locations, prior operations, indication). Once a classifier outputs a class (Ground 1), computing a confidence score and cross-checking it against patient metadata (a) is inherent to classifier design (probability outputs) and (b) flows directly from R1, which already ingests an acquisition context expressly including "the age" and a clinical/observation context. Combining a classifier's confidence score with the demographic/patient-file context data already taught by R1 is an obvious use of known data for its known purpose.

4.5 Corroborating context (caution — not § 103 prior art)

The Cited By table on the family's pre-grant publication (US 2018/0100907 A1) lists, among others:

• US 10,776,917 B2 (Siemens, priority 2017‑10‑27) — motion-artifact compensation by machine learning;
• WO 2019/201968 A1 / US 11,333,732 B2 (Koninklijke Philips) — "Automatic artifact detection and pulse sequence modification in magnetic resonance imaging";
• DE 10 2018 207 632 A1 / US 11,547,368 B2 (Siemens) — determining an imaging modality and its parameters;
• EP 3 597 107 A1 / US 11,925,501 B2 (Siemens) — topogram-based fat quantification.

These all post-date the 2016‑10‑12 priority and therefore cannot be used as § 102/103 prior art. I note them only because the Philips reference is, on its face, the same inventive concept (detect artifact → modify acquisition parameters) arriving from a different company within roughly two years. I would not rely on this, and it is not an objective-indicia argument — contemporaneous independent development by others is weak evidence at best. I flag it as a lead for locating true pre-2016 counterparts to R4/R5, which is where the strongest additional art would be found.

4.6 What I could not find

• No IPR, PGR, or district-court validity challenge to US 11,402,453 appears in the record I retrieved. The PTAB briefing that surfaced in searches concerned unrelated patents ('940, '373 — semiconductor/SQC subject matter) and is not relevant here.
• I could not retrieve the actual examiner-cited references or the applicant's IDS from this page. If the task intends a specific Prior Art list from the Google Patents "References Cited" table, that data was not present in the supplied text and the analysis above should be re-run against it.

5. Anticipated patent-owner rebuttals, and how they fare

A. "The prior art is silent on modifying acquisition parameters in response to a post-processing-capacity analysis."
True as to R1. This is the strongest non-obviousness argument, and it is why Ground 1 turns on the sufficiency of R5 + motivation. It fails under KSR if the reacquisition arm can be tied to a concrete, articulated reason — which the patent's own Background supplies (the recall problem) and which clinical practice supplies (re-scan while on table).

B. "R1 teaches away: its solution to unsuitable data is to choose a different tool, not to re-scan."
A colorable In re Gurley / In re Fulton argument. But teaching away requires that a POSITA reading the reference as a whole be led in a divergent direction — not merely that the reference chose one of two available remedies. R1 does not disparage reacquisition, and its stated goal (avoiding useless data) is served by reacquisition rather than defeated by it. Weak-to-moderate.

C. "The neural network trained on reference data labeled by post-processing outcome is a specific, non-obvious training methodology."
Weak. Labeling supervised training data by the outcome of the downstream process is the standard technique; moreover, the '453 specification's own disclosure of this "learning method" (Figure 4, steps i–v) is presented at a level of generality that reads as a restatement of routine supervised learning with an in-line retraining step. Claim 23 (link result image to learning data by sufficiency) is a textbook training-set-construction step.

D. Objective indicia.
None appear to be on the record. No evidence of commercial success, long-felt-but-unsolved need, failure of others, or industry praise was found. Anexamination of family prosecution files would be needed to know whether any was argued.

6. Bottom line

Claim 1 would likely be obvious over R1 (US 2007/0008172 A1, or its grant counterpart US 8,036,434 B2) in view of machine-learning medical image-quality/artifact-classification art and adaptive scan-parameter/reacquisition art, with the motivation supplied by (i) the identity of the problem addressed by R1, (ii) the patient-recall problem admitted in the '453 Background, (iii) the routine nature of the claimed control parameters (claim 9), and (iv) KSR's obvious-to-try and predictable-substitution principles. The sole element requiring a genuine motivation argument is the closed reacquisition loop of claims 1(f)–(h); every other element is either squarely in R1 or an admitted/routine practice.

Dependent claims 6, 9, 10, 11, 12, 13, 14, 19, 21, 22 and 24 add nothing patentably distinct: plausibility checks over demographic/patient-file data (cl. 6) use context data R1 already teaches; the parameter list (cl. 9) is conventional; iteration and real-time operation (cl. 10–12) recite logistics; topogram-based operation (cl. 13–14) is the purpose of a topogram; user selection of the desired process (cl. 19) and the enumerated post-processing operations (cl. 21) are the admitted prior-art catalogue of the specification's own Background. Claims 8 and 18 (patient remains in the modality) are the closest to a genuine structural difference, but they recite where the patient is, not what the system does differently.

Confidence: moderate. The mapping of R1 to elements (a)–(d) is well grounded in verified text. The R4/R5 legs rest on my assessment of the 2016 state of the art without verified reference numbers, because this page's Prior Art section was not present in the supplied material. Confirming the Examiner's cited art and the IDS from the file wrapper (or the Google Patents "References Cited"/"Cited By" tables) would likely move this from moderate to high confidence — and, given the density of the field, would likely add closer art than R1 on the neural-network and adaptive-protocol elements.